Generation of foxq1a ^bcz11 and foxq1b ^bcz18 null alleles using CRISPR-Cas9

We employed the highly effective CRISPR-Cas9 mediated targeted mutagenesis in zebrafish as previously described [53,54] to create loss-of-function mutations at the beginning of foxq1a and foxq1b genes. We isolated two mutant alleles of interest: foxq1a^bcz11 and foxq1b^bcz18 (Fig 1A). These are frameshift mutations that lead to a premature stop codon at the beginning of each respective gene: bcz11 has a large 95 base pair (bp) deletion in foxq1a that is immediately followed by a premature stop codon, and bcz18 has a 4 bp deletion and a 12 bp insertion in foxq1b that causes a premature stop codon within the start of the indel (Fig 1A). RT-PCR and sequencing analysis confirmed expression of the expected nonsense transcripts from the bcz11 and bcz18 mutations, and no alternative splicing was found (S1 Fig). The data indicate that bcz11 and bcz18 represent null alleles of their respective genes by producing nonsense mRNAs that would prevent Foxq1a and Foxq1b protein expression, respectively. For each gene target, three guide RNAs (gRNAs) were used in combination to increase the efficiency of the mutagenesis and the probability of generating a large indel mutation (Fig 1A). At 5 days post-fertilization (dpf) when zebrafish have well established organ systems and are free-swimming larvae, we examined the gross morphology of the homozygous mutants (Fig 1B). We found that both foxq1a^bcz11 and foxq1b^bcz18 homozygous mutants were indistinguishable from wildtype siblings (Fig 1B), displaying normal-sized inflated swim bladders, and grossly normal head and body structures that were no different from wildtype. However, we cannot exclude the possibility that craniofacial and specific organ phenotypes may be present in the mutants without further detailed analysis, in regions that normally express foxq1a and foxq1b [51,52]. The viability of bcz11 and bcz18 mutants, at least up to larval stages, appeared to be normal, as we consistently found ~25% of the progenies to be mutants from a heterozygous incross.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Sequence analysis of foxq1a^bcz11 and foxq1b^bcz18 mutations generated by CRISPR-Cas9 targeting.

a Top, DNA and amino acid sequences of the 5’ end of foxq1a coding region. DNA sequencing chromatograms show the expected foxq1a sequence in the wildtype sibling, and a large 95 base pair deletion (red box) in the homozygous bcz11 mutant, which causes a premature stop (black box with an asterisk) at the beginning of the gene. Bottom, DNA and amino acid sequences of the 5’ end of foxq1b coding region. DNA chromatograms show wildtype sequence in foxq1b wildtype sibling, while the homozygous bcz18 mutant carries a 4 bp deletion and 12 bp insertion (red box) causing a nonsense mutation at the beginning of the gene. gRNA, target sites of the guide RNAs used. b Whole mount live imaging of 5 dpf larvae show normal gross morphological development of foxq1a^bcz11 and foxq1b^bcz18 mutants as compared to wildtype siblings (at least 10 animals were analyzed per genotype group).

https://doi.org/10.1371/journal.pone.0194207.g001

Characterization of foxq1a and foxq1b gene expressions

The expression patterns of foxq1a and foxq1b in zebrafish remain largely unanalyzed. Only two studies to date have examined their expressions in specific tissues of interest; one showed foxq1b expression in the developing jaws at 2 dpf [51] and the other reported that foxq1a may be expressed in the cranial suture in juvenile zebrafish (6 weeks post-fertilization) [52]. To better define the possible functions of these genes in zebrafish, we expanded the characterization of their expressions to various stages of development from 1 to 8 dpf, and tested whether they were expressed in macrophages (Fig 2). Using fluorescence-activated cell sorting (FACS), we isolated macrophages (M) based on GFP expression from 2.5 dpf zebrafish embryos carrying the macrophage-specific transgene mpeg1:EGFP (Fig 2A). To assess their expression in other cell types as additional references, we sorted erythrocytes (E) based on DsRed expression from erythrocyte-specific gata1:DsRed transgenic embryos as well as the remaining non-fluorescent cells (A) (Fig 2B and 2C). We verified the identity of the sorted cell populations by RT-PCR analysis using known markers and fluorescent reporter genes for the E, M, and A cells (Fig 2B and 2C). The data shows that while both genes are expressed from 2 to 8 dpf (Fig 2C), only foxq1a is expressed by macrophages and neither gene is expressed by erythrocytes (Fig 2C). Using RNA in situ hybridization, we found expression of foxq1a in the cranial region at 1 dpf that persists in the jaw at 2.5 dpf when it is also found in the body tissue along the yolk, and on the yolk (S2 Fig). By contrast, foxq1b was primarily found to be expressed in the ventral jaw region at 2.5 dpf (S2 Fig). While foxq1a RNA expression was found in macrophages isolated from wildtype embryos (Fig 2C), its level may be too low to be detectable by whole mount RNA in situ hybridization as we were unable to visualize its expression in macrophages by in situ (S2 Fig). Because FOXQ1 is known to be highly expressed in the mammalian gastrointestinal tract [15,16,55,56], we examined whether foxq1a and foxq1b are expressed in the adult zebrafish gut, and indeed found them both to be expressed (Fig 2C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Gene expression analysis of foxq1a and foxq1b.

a Isolation of macrophages based on GFP expression by FACS in mpeg1:EGFP transgenic zebrafish embryos at 2.5 dpf. Non-fluorescent embryos were used as a negative control for gating. Top left, P4 shows the cell fraction sorted as GFP+ macrophages, also shown in the right panels in brightfield and green channel. b RT-PCR analysis of gene markers validated the different cell populations sorted by FACS, as denoted by E, erythrocytes; M, macrophages; A, all remaining non-fluorescent cells. The following genes were used: translation elongation factor 1 (ef1a) as a reference marker, hemoglobin beta embryonic-1.1 (hbbe1.1) as an erythrocyte marker, and macrophage expressed gene 1 (mpeg1) and interferon regulatory factor 8 (irf8) as well-established macrophage markers in zebrafish. As expected, the ‘A’ cells expressed all genes, while erythrocytes ‘E’ cells expressed hbbe1.1, ef1a, and DsRed genes and macrophage ‘M’ cells expressed macrophage markers, GFP, and ef1a. c Using the sorted cell populations, we found expression of foxq1a in macrophages and neither gene in erythrocytes. Both genes are expressed at 2 to 8 dpf of development as well as in the adult gut tissue.

https://doi.org/10.1371/journal.pone.0194207.g002

Single and double foxq1a^bcz11 and foxq1b^bcz18 mutants show normal macrophage, neutrophil, and microglia development

In light of the finding that foxq1a is expressed in macrophages and the implicated mammalian FOXQ1 function in tumor biology that may involve innate immune cells [19], we asked whether the development of macrophages and neutrophils may be altered in the zebrafish foxq1 mutants. By whole mount RNA in situ hybridization using a macrophage marker (mfap4) [57,58] and neutrophil marker (mpx) [59,60], we found no apparent difference in formation or distribution of these innate immune cells during development at 2.5 dpf in foxq1a^bcz11 mutants and foxq1b^bcz18 mutants (Fig 3A). In case of redundant or compensatory functions of foxq1a and foxq1b, we examined the double foxq1a^bcz11;foxq1b^bcz18 homozygous mutants, but also found no marked differences compared to wildtype siblings (Fig 3A). Quantification of macrophage and neutrophil numbers in the tail region further substantiates our finding that the mutants are not significantly different from siblings in innate immune cell development (Fig 3B). Since it is possible to have defects in macrophages at later time points, we evaluated the development of microglia, an important derivative of embryonic macrophages. We used a vital neutral red stain assay that has been shown to specifically label microglia in the larval zebrafish brain [61–63] as cells with dark red aggregates (Fig 4A). We found that all foxq1a^bcz11, foxq1b^bcz18 and foxq1a^bcz11;foxq1b^bcz18 mutants exhibited the typical pattern (Fig 4A) and numbers (Fig 4B) of microglia in the larval midbrain as those in wildtype and heterozygous control siblings and previously described for wildtype [61–65]. Taken together, the data indicates that foxq1a and foxq1b are not required for the proper development of macrophages, neutrophils, and microglia.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Single and double foxq1a^bcz11 and foxq1b^bcz18 mutants have normal macrophage and neutrophil development.

a Whole mount in situ hybridization of 2.5 dpf single and double foxq1a^bcz11 and foxq1b^bcz18 mutants and their siblings using RNA probes for mfap4 (macrophage marker) and mpx (neutrophil marker). Arrows, macrophages or neutrophils in the caudal hematopoietic tissue in the embryo tail. b Scatter plots showing number of macrophages and neutrophils in the tail region for each embryo quantified in the different genotype categories. n, number of embryos analyzed beneath each scatter plot. Plots report average ± standard deviation (s.d.). Multiple unpaired t-tests comparing between control siblings and each of the mutants groups were conducted with correction for multiple comparisons using the Sidak-Bonferroni method and without assuming equal variance to determine statistical significance. n.s., no statistical significance as defined by p > 0.05 was found in all the comparisons.

https://doi.org/10.1371/journal.pone.0194207.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Neutral red analysis shows normal microglia development in single and double foxq1a and foxq1b mutants at 4 dpf.

a Neutral red staining in 4 dpf wildtype (WT) sibling, foxq1a^bcz11, foxq1b^bcz18, and foxq1a^bcz11;foxq1b^bcz18 mutants shows a stereotypical pattern of microglia in the midbrain (arrow). b Bar graph shows all larvae analyzed in all genotype groups had normal numbers of microglia (>25). Number of larvae analyzed (n) shown to the right of the corresponding bar.

https://doi.org/10.1371/journal.pone.0194207.g004

foxq1a and foxq1b null mutants exhibit typical transcriptional response to Escherichia coli brain challenge

While foxq1a and foxq1b are not essential for innate immune cell development, they may affect the function of these cells, such as their ability to mediate an immune response. We developed a brain challenge assay (Fig 5A) to trigger innate immune activation by introducing live E. coli bacteria into the brain tectum of zebrafish larvae. We used this assay to address whether foxq1a, foxq1b, or both may regulate systemic response to E. coli challenge. We validated the effectiveness of our assay to trigger immune activation by showing that after brain challenge with bacteria or LPS at 6 hours post injection (hpi), we indeed found significant elevation of several innate immune activation genes systemically: interleukin 1 beta (il1β) [61,62,66], immune responsive gene (irg1) [67], neutrophil gene (mpx) [68], and tumor necrosis factor alpha (tnfα) [69] but not in water injected or uninjected controls (Fig 5B). We also tested whether foxq1a and foxq1b themselves could be transcriptionally regulated during a bacterial response to implicate a possible role in this process. Interestingly, we found that foxq1a, but not foxq1b, was significantly downregulated by about a factor of 2 in wildtype larvae after LPS or bacteria injection (Fig 5B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. foxq1a^bcz11 and foxq1b^bcz18 mutants exhibit wildtype transcriptional response to E. coli challenge in the larval brain.

a Graphical representation of the brain challenge assay. E. coli was injected into right brain tectum, and at 6 hours post injection (hpi), injected larvae were individually processed for RNA extraction, genotyping, cDNA synthesis, and qPCR analysis. b Bar chart shows significant upregulation of immune activation genes il1β, irg1, mpx, and tnfα after bacteria challenge at 6 hpi in wildtype larvae compared with uninjected controls. Water injected wildtype larvae exhibit no induction of activation genes, whereas LPS injected wildtype animals have a significant upregulation of il1β and irg1. After microbial activation by E. coli or LPS, a significant downregulation of foxq1a was found by about a factor of 2. No significant change was found for foxq1b after immune challenge. Data from wildtype injections validate the efficacy of the assay to activate the innate immune system. Dotted line marks no fold difference at 1. At least 6 or more independent biological samples were measured per category. c Bar plot shows transcriptional changes of target genes after E. coli injection in the brain at 6 hpi for control siblings and foxq1a and foxq1b mutants. n = 9–14 independent biological samples were measured per genotype. All error bars show standard error of means. Statistical significance in b and c was determined by multiple unpaired t-tests comparing between uninjected or sibling controls and the experimental groups with correction for multiple comparisons using the Sidak-Bonferroni method and without assuming equal variance. n.s., no statistical significance as defined by p > 0.05. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

https://doi.org/10.1371/journal.pone.0194207.g005

To analyze the immune response of foxq1 mutants, we employed the brain challenge assay on 4 dpf larval progenies of foxq1a^bcz11/+ and foxq1b^bcz18/+ single and double heterozygous incrosses (Fig 5C). We found that the expression levels of immune activation genes (il1β, irg1,mpx, and tnfα) in foxq1 mutants at 6 hpi were not significantly different from WT siblings (Fig 5C). However, we cannot exclude the possibility that double foxq1a and foxq1b mutants may have deficient or aberrant transcriptional response to E. coli. Since foxq1a transcriptional level was downregulated in wildtype response to E. coli (Fig 5B), we also assessed whether this occurred normally in foxq1 mutants as compared to WT siblings. qPCR data shows that foxq1b mutants had similar foxq1a expression as WT controls after bacterial exposure, but both heterozygous and homozygous foxq1a mutants had a modest but significant increase of ~50% in foxq1a RNA level (Fig 5C). The foxq1a upregulation was also found in foxq1a mutants at steady state, suggesting that the increase during the bacterial response may be due to an already higher baseline level rather than necessarily the bacterial exposure (S3 Fig). The results indicate that foxq1a and foxq1b may not be required for mediating an appropriate transcriptional response to E. coli challenge, but foxq1a may affect other aspects of this bacterial response as it is transcriptionally regulated.

Zebrafish lines and embryos

Embryos from wildtype (TL and AB), transgenic lines: mpeg1:EGFP [58,71] and gata1:DsRed [72], and heterozygous lines: foxq1a^bcz11/+, foxq1b^bcz18/+, and foxq1a^bcz11/+ foxq1b^bcz18/+ were raised at 28.5°C, and staged as described [73]. Homozygous mutants were derived from either single heterozygous incrosses or from double heterozygous incrosses. The latter yielded double foxq1a^bcz11 foxq1b^bcz18 mutants at a ratio of ~ 1/16. Embryos were treated with 0.003% 1-phentyl-2-thiourea (PTU) to inhibit pigmentation in sterilized zebrafish system water. This study was carried out in accordance with the animal protocol (16–160) approved by the Institute Animal Care and Use Committee at UNC Chapel Hill.

Neutral red analysis

Microglia were scored in live larvae by neutral red vital dye staining assay as previously described [61,62]. In brief, 3 dpf larvae were stained with neutral red by immersion in system water supplemented with 2.5 μg/ml neutral red and 0.003% PTU at 28.5°C for 1 hour, followed by 1–2 water changes, and then analyzed 2–3 hours later using a stereomicroscope.

Whole mount RNA in situ hybridization

In situ hybridization was performed using standard methods as described in [61,62,74]. Antisense riboprobes were transcribed from: mfap4 plasmid previously cloned [61], mpx (full-length, Open Biosystems clone 6960294) [61], foxq1a plasmid (pCES150), which was made from a 877 bp fragment cloned from a wildtype zebrafish embryo cDNA library using primers F476: GCACTCATCATCTGCAACAGGTA and R1353: TGATATCCCGCGGTTGCAGG. Anti-sense foxq1a RNA probe was made from pCES150 digested with HindIII and transcribed by T7 RNA polymerase. To make the negative control RNA probe, foxq1a sense strand sequence was transcribed by T3 RNA polymerase using a PCR template made from the foxq1a plasmid pCES150. foxq1b anti-sense RNA probe was made from a PCR-based template encompassing 754 bp of coding sequence using forward primer: GTGTTGCGAGATCCCTCGCGTC and reverse primer: GTAAACACTGTGCAGTGGCGCGTC, where the reverse primer also included the T7 RNA polymerase recognition site.

RNA extraction, RT-PCR and quantitative PCR (qPCR)

Total RNA was extracted from individual larvae using the RNAqueous-Micro Isolation Kit (Ambion). cDNA was made using oligo dT primer and SuperScript IV reverse transcriptase (Invitrogen). qPCR was performed on the Applied Biosystems QuantStudio Flex 6 Real-Time PCR System using probe-based assays or SYBR Green. The delta–delta ct method was used to determine the relative levels of mRNA expression between experimental samples and controls. ef1a was used as the reference gene for determination of relative expression of all target genes. Sequences of the qPCR probes and primer sets used in qPCR and RT-PCR analyses are listed in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Probes and primer sequences for qPCR and RT-PCR.

https://doi.org/10.1371/journal.pone.0194207.t001

CRISPR-Cas9 targeted mutagenesis of foxq1a and foxq1b

The target genes were foxq1a (NCBI accession: NM_001243344.1; Gene ID: 100537750) and foxq1b (NCBI accession: NM_212907.1; Gene ID: 405843). Co-injection of Cas9 mRNA and guide RNAs (gRNAs) was conducted in wildtype 1-cell stage zebrafish embryos. Cas9 mRNA was transcribed from XbaI linearalized pT3TS-nCas9n plasmid (Addgene #46757) [53] using mMessage mMachine T3 Kit (Ambion) according to the manufacturer’s instructions. CRISPR targets for gRNA designs were identified using CHOPCHOP (http://chopchop.cbu.uib.no) [54]. The following gene-specific oligonucleotides using T7 promoter were used to make gRNAs as previously described [54]; gene-specific target sequences are underlined: foxq1a gRNA-1 5’-TAATACGACTCACTATAGGGTTCGAAGAGTGACAGGGGTTTTAGAGCTAGAAATAGCAAG-3’, foxq1a gRNA-2 5’-TAATACGACTCACTATAGGACTACGACTCCAAGCCTGGTTTTAGAGCTAGAAATAGCAAG-3’, foxq1a gRNA-3 5’- TAATACGACTCACTATAGGAGTTGTGCAGCGATGCTGGTTTTAGAGCTAGAAATAGCAAG-3’, foxq1b gRNA-1 5’-TAATACGACTCACTATAGGCTTCACCGCTGTCCACGGGTTTTAGAGCTAGAAATAGCAAG-3’, foxq1b gRNA-2 5’-TAATACGACTCACTATAGGTTCCTCCTCCGTGGACAGGTTTTAGAGCTAGAAATAGCAAG-3’, foxq1b gRNA-3 5’-TAATACGACTCACTATAGGAGCCCAATTCCTCCTCCGGTTTTAGAGCTAGAAATAGCAAG-3’. In vitro transcription of gRNAs from assembled oligonucleotides was conducted using the HiScribe T7 Quick High Yield RNA Synthesis Kit (NEB). Injected clutches of embryos validated to contain CRISPR mediated mutagenesis by a T7 endonuclease assay were raised as F0 fish. Incross and outcross of F0 adults produced F1 progenies, which were genotyped and sequenced at adult stages to identify F1 founders carrying mutations of interest. These founders were isolated and outcrossed to raise new mutant lines, which yielded the foxq1a^bcz11/+ and foxq1b^bcz18/+ heterozygous fish used in this study.

Cell dissociation and fluorescence activated cell sorting (FACS)

Dissociation of pools of 25–50 2.5 dpf transgenic zebrafish embryos either carrying the macrophage reporter (mpeg1:EGFP) or the erythrocyte reporter (gata1:DsRed) was conducted using a protocol adapted from [75]. Fluorescently tagged cells were sorted from non-fluorescent cells using the BD FACSAria cell sorter equipped with three lasers. Non-fluorescent control embryos were used for setting gates in FACS for GFP and DsRed positive cells. BD FACSDiva V8.0.1 software was used for cell sorting and FACS plots. Total RNA was extracted from the different populations of cells (GFP positive, GFP negative, DsRed positive, DsRed negative) using the RNAqueous-Micro Isolation Kit (Ambion), and made into cDNA for RT-PCR analyses as described above.

Genotyping bcz11 and bcz18 alleles

To genotype foxq1a^bcz11, the following primers were used: foxq1a F 5’-CGGTGTTTTTGTGACTTGATTT-3’ and foxq1a R 5’-AAGAGTATGGAGGTTTGGGTCTC-3’ in a PCR assay that yielded differentially sized products to distinguish the mutant allele which has a 95 bp deletion compared with the wildtype allele. The expected sizes for mutant band is 240 bp and wildtype band is 335 bp. To genotype foxq1b^bcz18, a PCR and restriction enzyme combined assay was used. Primers foxq1b F 5’-AAGCAACTCATCTGACCTGACA-3’ and foxq1b R 5’-GGGTCTACGCGTATATGGTTTC-3’ were used to yield PCR products of 217 bp, followed by a digest with BsaJI, which only recognizes the wildtype sequence.

Immune challenge

Bacteria were prepared from 3 mL overnight cultures, each derived from a single Escherichia coli colony, which were centrifuged and re-suspended in 500 μL of PBS. The 500 μL bacteria mix (~1.6 x 10⁶ cfu/μL) was supplemented with 1 μL of 5 mg/ml AlexaFluor 568 or AlexaFluor 488 conjugated dextran (10 kDa) from Invitrogen. Other reagents used for brain injections were: ultra-pure water supplemented with 0.05 mg/ml fluorescently labeled dextran (10 kDa) for the control water samples, and lipopolysaccharides (LPS) derived from 0111:B4 E. coli (L3024 Sigma) at 4.5 mg/ml was supplemented with 0.05 mg/ml fluorescently labeled dextran (10 kDa). For all injections, 1 nL of the reagent mix was injected into the right brain tectum of 4 dpf larvae, and the fluorescent dextran was used to validate the injection site. Larvae that were successfully injected were collected 6 hours post injection (hpi) and immediately processed for total RNA lysis. RNA extraction and subsequent steps for qPCR analyses are described above. All collected larvae derived from heterozygous intercrosses were genotyped after RNA extraction.